EP2895834A1 - Capteur tactile conforme doté d'un matériau de type éponge rempli de fluide - Google Patents

Capteur tactile conforme doté d'un matériau de type éponge rempli de fluide

Info

Publication number
EP2895834A1
EP2895834A1 EP13766791.1A EP13766791A EP2895834A1 EP 2895834 A1 EP2895834 A1 EP 2895834A1 EP 13766791 A EP13766791 A EP 13766791A EP 2895834 A1 EP2895834 A1 EP 2895834A1
Authority
EP
European Patent Office
Prior art keywords
tactile sensor
fluid
sponge
compliant tactile
compliant
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
EP13766791.1A
Other languages
German (de)
English (en)
Inventor
Jeremy A. FISHEL
Gerald E. Loeb
Raymond A. Peck
Chia-Hsien Lin
Blaine MATULEVICH
Vikram PANDIT
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
SynTouch LLC
Original Assignee
SynTouch LLC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by SynTouch LLC filed Critical SynTouch LLC
Publication of EP2895834A1 publication Critical patent/EP2895834A1/fr
Withdrawn legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01L—MEASURING FORCE, STRESS, TORQUE, WORK, MECHANICAL POWER, MECHANICAL EFFICIENCY, OR FLUID PRESSURE
    • G01L1/00—Measuring force or stress, in general
    • G01L1/02—Measuring force or stress, in general by hydraulic or pneumatic means
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01L—MEASURING FORCE, STRESS, TORQUE, WORK, MECHANICAL POWER, MECHANICAL EFFICIENCY, OR FLUID PRESSURE
    • G01L5/00—Apparatus for, or methods of, measuring force, work, mechanical power, or torque, specially adapted for specific purposes
    • G01L5/22—Apparatus for, or methods of, measuring force, work, mechanical power, or torque, specially adapted for specific purposes for measuring the force applied to control members, e.g. control members of vehicles, triggers
    • G01L5/226—Apparatus for, or methods of, measuring force, work, mechanical power, or torque, specially adapted for specific purposes for measuring the force applied to control members, e.g. control members of vehicles, triggers to manipulators, e.g. the force due to gripping
    • G01L5/228—Apparatus for, or methods of, measuring force, work, mechanical power, or torque, specially adapted for specific purposes for measuring the force applied to control members, e.g. control members of vehicles, triggers to manipulators, e.g. the force due to gripping using tactile array force sensors

Definitions

  • This disclosure relates to robotics and prosthetic limbs, including fingertips.
  • Robots may be unable to carry on a broad array of important delicate tasks, such as assembly operations, handling fragile objects, and preventing damage in accidental collisions.
  • Prosthetics may similarly be unable to carry on a broad array of important delicate tasks.
  • Robotic systems may include a rigid set of links with actuated degrees of freedom that are controlled with electric motors.
  • the position, velocity, and/or force at these degrees of freedom may be controlled by a high-level controller coordinating the movement of each of the joints.
  • the controller may run completely autonomously or be controlled manually in whole or in part by a human operator, as in the case of telerobotics.
  • Such robotic systems may physically interact with their environment.
  • the high-level controller may obtain knowledge about its environment and respond to both expected and unexpected events.
  • machine vision may be employed to identify the location and orientation of objects and precision equipment may direct the robot to a desired location.
  • algorithms for machine vision may be slow and prone to errors when vision is obstructed, surfaces are out of view, or there are shadows or is poor contrast.
  • a human operator may interpret images from a video camera or observe the robot directly and may attempt to adjust his or her commands to the robot as rapidly and precisely as possible.
  • this mental concentration may be exhausting to the operator and reaction time may be slow.
  • Fast-reacting systems may require precise robotics employing stiff mechanical linkages, high-quality position and force encoders, high-speed feedback controllers, and powerful and heavy motors.
  • DC motors may drive industrial robotic hands or myoelectric prosthetic hands.
  • the closing speed of an unloaded hand may be directly proportional to the voltage across the motors.
  • this control voltage may be generated from amplified and filtered electromyography recordings of a prosthesis user's residual muscles to produce a voltage in proportion to muscle activation strength to capture a user's intent for opening or closing a hand. See Sears HH, Shaperman J., "Proportional myoelectric hand control: an evaluation," Am J Phys Med Rehabil. 1991 Feb, 70(1 ): 20-8; Kyberd PJ, Chappell PH.
  • Fragile or compliant objects may therefore be challenging to grasp without damage or deformation because they may not be capable of resisting such forces. If the operator decides to pick up a fragile object, therefore, he or she may need to use a high level of visual attention to precisely time stopping the hand as it encloses on an object before it breaks. Such a task can be quite difficult with prosthetic hands due to delays in visual processing in the cortex (e.g., ⁇ 200ms), filtering delays in EMG signals, and inertia and friction of motors. This can make timing a precise stopping of fingertips quite challenging.
  • a similar control problem may arise in preventing robot appendages from harming themselves or damaging external objects when collisions occur with or between appendages.
  • These appendages may have hard surfaces.
  • the impact force from a collision with a robotic actuator may increase with the hardness of the colliding surfaces and the relative speed of the robotic appendage and the object.
  • Machine vision in particular can be subject to errors if the cameras become occluded or if lighting is poor. In applications where safety and reliability are of high concern, these technologies may therefore be less desirable.
  • BioTac® For prosthetic and robotic fingers, there is a commercially-available product called a BioTac® in which contact is sensed by a pressure sensor connected to a liquid that is used to inflate an elastomeric skin over a rigid core. See US Patent Numbers: 76581 19, 7878075, 8181540, 8401658, and Fishel JA, Santos VJ, Loeb GE. "A robust micro-vibration sensor for biomimetic fingertips," IEEE/BioRob. IEEE, 2008, pp. 659-63; Wettels N, Santos VJ, Johansson RS, Loeb GE. "Biomimetic tactile sensor array " Adv. Robotics.
  • the device has a compliance similar to the human fingertip and can provide sensitivity that exceeds human performance. See Fishel JA, Loeb GE. "Sensing tactile microvibrations with the BioTac— Comparison with human sensitivity," IEEE/BioRob. IEEE, 2012b, pp. 1 122-7.
  • the device has a complex electromechanical design and may require an electrically conductive and incompressible liquid, typically a form of saltwater. This may damage mechatronic components if it leaks from the sensor. Further, differences between the inertial properties of the liquid-filled fingertip and the ambient air may amplify vibration from motor actuation so as to create background noise in the pressure
  • the BioTac and other inflated sensors may require and result in a skin surface that is convex as a result of inflating an elastic skin with a fluid material.
  • the shape of an appendage that requires tactile sensing may include complex and compound curves, including regions that are concave, regions that must be subdivided into separately sensed compartments, and regions that project over structures to protect them, but cannot attach to those structures without interfering with their motion. If the fluid inflating the elastic skin escapes, moreover, the skin may lose its desired shape and may collapse onto the underlying rigid structure, thereby losing compliant protection of that rigid structure.
  • Tactile sensors may detect a wide range of physical phenomena, including capacitive, optical, magnetic, inductive, resistive, piezoelectric, piezoresistive, and ultrasonic. See Nicholls HR, Lee MH. "A survey of robot tactile sensing technology ' ' Intnl J Robotics Res. 1989, 8(3): 3-30; Howe RD. "Tactile sensing and control of robotic manipulation,” Adv. Robotics. 1994, 8(3): 245-61 ; Lee MH, Nicholls HR. "Tactile sensing for mechatronics - a state of the art survey ", Mechatronics. 1999, 9:1-31 ; and Dahiya RS, Metta G, Valle M, Sandini G. "Tactile sensing— from humans to humanoids," IEEE Trans Robotics. 2010, 26(1 ): 1-20.
  • Some tactile sensing approaches have sensor arrays that do not offer much compliance and tend to be insensitive to contact forces applied between discrete tactile cells. Examples of these include the Takktile array (Takktile, LLC), RoboTouch (Pressure Profile Systems), Weiss Tactile Sensors (Weiss Robotics) and the RoboSkin project. Patterning these sensory cells over complex surface can be challenging and costly and may result in areas that are insensitive to contact at the areas where contact sensitivity becomes the most important, such as edges and joints.
  • a compliant tactile sensor may include sponge-like material, a flexible skin, and a fluid pressure sensor.
  • the flexible skin may have a shape, absorb fluid, compress in response to force applied to the sponge-like material, and decompress and return to its original shape when the force is removed.
  • the flexible skin may cover an outer surface of the sponge-like material.
  • the fluid pressure sensor may sense changes in pressure in fluid that is within the spongelike material caused by a force applied to the flexible skin.
  • a fenestration may connect the sponge-like material to the fluid pressure sensor and may allow fluid to flow between the sponge-like material to the fluid pressure sensor.
  • the sponge-like material may be an open-cell foam.
  • the flexible skin may be made of the same material as the foam and may be integral to the sponge-like material.
  • the sponge-like material may be an engineered material that has a repeating pattern of cells.
  • the flexible skin may constitute a coating on the sponge-like material.
  • the flexible skin may be impermeable to the fluid.
  • the sponge-like material and the flexible skin may both be permeable to the fluid, but the permeability of the flexible skin to the fluid may be substantially less than the permeability of the sponge-like material to the fluid.
  • Fluid may be within the sponge-like material.
  • the fluid may be a liquid or a gas.
  • the compliant tactile sensor may include a valve that, when open, allows fluid to flow between outside of the compliant tactile sensor and the sponge-like material.
  • the valve when open, may only allow the fluid to flow in one direction.
  • the valve may allow fluid to flow into the sponge-like material from outside of the compliant tactile sensor only when the pressure of the fluid outside of the compliant tactile sensor exceeds the pressure of the fluid within the spongelike material by a threshold amount.
  • the valve may allow fluid to flow outside of the compliant tactile sensor from within the sponge-like material only when the pressure of the fluid within the sponge-like material exceeds the pressure of the fluid outside of the compliant tactile sensor by a threshold amount.
  • a robotic system may include a movable robotic arm, a compliant tactile sensor on the movable robotic arm that senses contact between the compliant tactile sensor and an object during movement of the movable robotic arm and that cushions the effect of that contact, and a reflex system that causes the moveable robotic arm to move in response to commands.
  • the robotic system may include multiple compliant tactile sensors on the movable robotic arm that each sense contact between the compliant tactile sensor and an object during movement of the movable robotic arm and that each cushion the effect of that contact.
  • the robotic system may include a flexible joint and the compliant tactile sensor may cover the flexible joint.
  • the robotic system may include a controller that generates the commands and determines whether the movable robotic arm has come in contact with an object during movement of the movable robotic arm based on one or more signals from the compliant tactile sensor.
  • the controller may determine a magnitude of contact force that is applied to the compliant tactile sensor by the object based on one or more signals from the compliant tactile sensor.
  • the controller may stop movement of the movable robotic arm in the direction of the object when the one or more signals from the compliant tactile sensor indicate contact between the compliant tactile sensor and an object during movement of the movable robotic arm.
  • the robotic system may include a factor that provides tactile feedback of contact between the compliant tactile sensor and the object during movement of the movable robotic arm based on one or more signals from the compliant tactile sensor.
  • FIG. 1 illustrates an example of a compliant tactile sensor with ambient fluid, an elastomeric pad, and a sensor mount.
  • FIG. 2 illustrates an example of a compliant tactile sensor that has a check valve and vent for fluid.
  • FIGS. 3A-3C illustrate examples of robotic systems that have multiple compliant tactile sensors deployed on movable robotic arms.
  • FIG. 3A is an example of multiple compliant tactile sensors on a movable robotic arm that has a flexible joint
  • FIG. 3B is a cross-section of an example of multiple compliant tactile sensors on a movable robotic limb
  • FIG. 3C is an example of multiple compliant tactile sensors covering a flexible joint in a movable robotic arm.
  • FIG. 4 is an example of a robotic system that may include a movable robotic arm, a compliant tactile sensor, and various subsystems.
  • FIGS. 5A-5C illustrate an example of a force been applied to and then removed from two different types of compliant tactile sensors.
  • FIG. 5A illustrates the force
  • FIG. 5B illustrates the responses from the sensors
  • FIG. 5C illustrates a high-pass filtered version of the signals illustrated in FIG. 5B.
  • FIG. 6 illustrates an example of a rigid object colliding with movable robotic arm, both with and without an elastic skin covering.
  • FIGS. 7A-7C illustrate an example of how a compliant tactile sensor can mitigate contact forces.
  • FIG. 7A illustrates velocity during collision and loss of contact with a movable robotic arm both with and without a compliant tactile sensors;
  • FIG. 7B illustrates displacement; and
  • FIG. 7C illustrates force.
  • FIG. 8 is an example of orientation adjustments that may be made to a movable robotic arm that has multiple compliant tactile sensors.
  • Various compliant tactile sensors and associated robotic systems are now described. They may provide a sensitive yet robust means to sense contact between a robotic or prosthetic appendage and other objects.
  • the design and electrical connections may be very simple. Contact surfaces on a robotic or prosthetic appendage may be compliant, rather than hard. There may be sensorized contact surfaces that have arbitrary shapes.
  • the compliant tactile sensor may be produced in arbitrary shapes at low per-unit costs.
  • the compliant tactile sensor may provide a high degree of sensitivity to detecting contact, while a low incidence of false-positive signals from inertial acceleration of the sensor.
  • the compliant tactile sensor may provide high sensitivity to both the onset and offset of contact. The sensor may move between different environments (such as air and water) while maintaining sensitivity to contact.
  • the compliant tactile sensor may be incorporated into a system that interprets data from the sensor to minimize collision forces and improve safety in the event of accidental collision.
  • the compliant tactile sensor may be incorporated into a system that interprets data from it to facilitate robotic grasp of fragile objects.
  • the robotic system may be low in cost and provide only limited accuracy in its movements, but employ tactile information to compensate for delays and errors in its commands.
  • Such compliant tactile sensors and associated systems may alleviate or reduce cognitive burdens experienced by prosthetic hand users and other teleoperators when grasping fragile or compliant objects.
  • Such sensors and system may be employed with prosthetic hand users and other teleoperators and facilitate conscious perception of sensory events.
  • Prosthetic limbs and parts thereof may be in the form of a robot. When directly controlled by the operator wearing them, they are considered to be a telerobot. Any portion of the robot that can move through space is an appendage or moveable arm of the robot and any portion of the surface of the appendage or movable arm may come into contact with external surfaces, tools and objects, with itself, with human beings (including the operator), and/or with other appendages of the same or a different robot. All of these are included in the concept of "object contact.”
  • a compliant tactile sensor 1 may include an elastomeric pad 10 and a sensor mount 20.
  • the elastomeric pad 10 may include sponge-like material 12 containing fluid 13 surrounded by a flexible skin 14.
  • the sponge-like material 12 may be bonded 15 to a rigid core 22, which may be part of the sensor mount 20.
  • the sensor mount 20 may include a compression plate 24 which may push a pressure sensor housing 28 containing a pressure sensor 29 against a gasket 25 when tighteners 26 are installed and tightened, thereby forming a complete seal of fluid 13 in the elastomeric pad 10.
  • Pressure, force or vibration applied anywhere on the outside of skin 14 may cause a change in fluid pressure to propagate through the fluid 13 and through a fenestration 16 in sponge-like material 12 into pressure sensor housing 28 where it may be sensed by the pressure sensor 29.
  • the mechanical compliance of the elastomeric pad 10 may offer a cushioning from impact that can reduce impact force, while
  • a number of off-the-shelf technologies may be used for sensing pressure. They may incorporate both the pressure sensor housing 28 and the pressure sensor 29 in a single package.
  • An example is the MS 1471
  • Measurement Specialties which is a low-cost, low-pressure, miniature gage pressure sensor based on MEMS (microelectromechanical systems) that provides analog millivolt output in response to applied pressure.
  • Analog pressure sensors may facilitate custom signal processing; alternatively, off-the-shelf digital pressure sensors may provide a smaller form factor and a reduced cost.
  • the sponge-like material 12 may be an open-cell reticulated foam.
  • the skin 14 may form spontaneously on the surfaces of the foam where it contacts an injection mold in which it is formed, as in the case of a self-skinning foam.
  • the foam material for the elastomeric pad 10 may be of any type, such as
  • the flexible skin 14 can be produced over the sponge-like material 12 through spray coating, dip coating, or other means to create an airtight seal.
  • the coating materials may be of any type, such as polyvinyl fluoride, polyurethane, or silicone elastomers.
  • the mold may be lined with a sprayed-on or preformed skin 14 before injecting the foam material to form the sponge-like material 1 2 within it.
  • the fenestration 1 6 may be produced by puncturing or drilling the self- skinning foam used to produce the sponge-like material 12 and the skin 14.
  • the sponge-like material 12 may be produced with 3D printing that constructs an elastomeric matrix that can be of any variant of lattice shapes that are permeable to fluid, such as cubic or honeycomb.
  • the sponge-like material 12 may be formed by fusing together elastic beads with heat or glue in such a manner as to leave inter-connected, fluid-permeable spaces between them.
  • the fluid 1 3 may be a liquid or a gas.
  • the fluid 1 3 may be the same as fluid in the environment in which the sensor is operating. For example, when the sensor is used in air, the fluid 1 3 may be the air. When the sensor is used in water, the fluid 1 3 may be the water.
  • the use of ambient fluid for the fluid 1 3 may mitigate any complications that otherwise might arise due to diffusion or leakage of fluid into the environment from the compliant tactile sensor or vice versa.
  • the use of ambient fluid for the fluid 1 3 may minimize any mismatch between the inertial properties of the sensor and the ambient environment, thereby reducing the amplitude of vibrations that may be induced in the sensor not arising from contact with the skin 14.
  • the gasket 25 may be made of the same self-skinning foam used to produce the sponge-like material 12 and the skin 14. Alternatively, the gasket 25 may be an o-ring or similar molded rubber part, or any other type of gasket.
  • the rigid core 22 may be machined from a suitable material, such as plastic or metal, including aluminum, delrin, ABS plastic or similar, or injection molded, photo-polymerized, or laser sintered from a suitable precursor material.
  • the surface of the core may be roughened by bead-blasting, chemical etching, or similar processes to improve the adhesion of the foam to the core, which can be molded directly over the core during production.
  • the tighteners 26 can be a screw or other similar fastener or, for injection-molded plastics, a snap-fitting.
  • the pressure sensor housing 28 may be permanently affixed over the fenestration 16 in the elastomeric pad 10 by means of an adhesive, such as epoxy,
  • cyanoacrylate or a chemical weld, such as by a solvent.
  • a modification can be made to the sensor to include one or more valves 50 in order to control the fluid pressure, exhaust, and/or refill the elastomeric pad 10, including unidirectional check valves, bidirectional check valves, and others.
  • a polymeric duckbill valve such as the DU 027.001 -150.01 from MiniValve, or similar check valve may be used.
  • the valves 50 may be in line with vents 51 that access either fluid in the environment at ambient pressures or positively or negatively pressurize fluid. The use of either positive or negative pressured fluid connected to the vents 51 may permit for detection of leaks, if the skin 14 of the elastomeric pad 10 or the gasket 25 of the sensor mount 20 is ruptured or damaged.
  • Positively pressurized fluid may be used to inflate the sensor to a greater volume, improving the compressible distance of the elastomeric pad 10.
  • the valves 50 may be incorporated into the sensor so that fluid expressed from the sensor can be replaced by ambient fluid, which may be sucked in by negative pressure created by expansion of the sponge-like material 12.
  • the internal fluid of the elastomeric pad 10 can be maintained by slow leaks or diffusion through the skin 14.
  • FIG. 3A illustrates a robot 40 consisting of a robot base 43 with one or more limb segments 44 linked by articulations 41 that are axially revolute joints.
  • Multiple tactile sensors 1 may be attached to various locations on limb segments 44.
  • the elastomeric pad 10 of at least some tactile sensors 1 may be shaped so as to protrude over articulations 41 without interfering with the motion of articulations 41 .
  • FIG. 3B is a cross-sectional view that illustrates the use of multiple tactile sensors 1 disposed over the surface of one limb segment 44, thereby providing spatial information about point of contact during collisions with external objects.
  • FIG. 3C illustrates robotic limb segments 44 connected by hinge-like articulation 41 in which multiple tactile sensors 1 have elastomeric pads
  • the timing, time-course, amplitude, frequency spectrum, and/or other features of signals detected by and transmitted from the pressure sensor 29 may be combined with information about the movements of the robot to command changes in its movements or to aid in the localization or identification of external objects. Examples are provided in the following publications: Su Z, Fishel JA, Yamamoto T, Loeb GE. "Use of tactile feedback to control exploratory movements to characterize object compliance," Front. Neurorobot. 2012, 6; Fishel JA, Loeb GE. "Bayesian exploration for intelligent identification of textures," Front.
  • FIG. 4 is an example of a generic control system making use of sensory information from tactile sensors 1 mounted on a robot 40.
  • a controller 2 may be a computer running software autonomously, or a human operator controlling a telerobotic system or prosthetic hand that sends motor commands to a reflex system 3, which may modify the motor commands going to one or more robotic actuators 42.
  • Reflex system 3 may be composed any type of signal processing electronics and firmware designed to receive and process electrical signals from the controller 2 intended to drive the one or more robotic actuators 42 that may or may not modify these electrical signals to one or more of the robotic actuators 42, based on sensory information measured from one or more of the tactile sensors 1 .
  • the function of the controller 2 and the reflex system 3 could be integrated into a single piece of hardware.
  • the robotic actuators 42 could be any type of actuator capable of changing the position of or applying torque or force through articulation 41 , including a direct drive electrical, pneumatic or hydraulic motor, a gear or cable mechanical linkage, or any other type.
  • Information obtained from the tactile sensors 1 mounted on any of the limb segments 44 of the robot 40 can be processed by a signal processor 6 and delivered to the reflex system 3 to modify the motor commands delivered from the controller 2 to the actuators 42 of the robot 40 to produce useful functions as described below.
  • Algorithms embodying such reflexive behaviors may mimic biological reflexes in the spinal cord, which may be modulated by descending commands from the brain to achieve useful functions for humans to enable stable grasp and protect the human body from damage.
  • movement commands may originate from autonomous software and, in the case of telerobotics movement, commands may originate from a human operator.
  • Signal processing in the reflex system 3 may utilize analog electronic circuitry to modify high power signals that drive the motors to move robotic actuators.
  • reflex system 3 can be directly integrated into higher-level analog or digital systems to modify command signals before they are amplified to the higher-power signals typically used to drive electric actuators.
  • Information derived from ambient fluid pressure can also be provided directly to an operator of a telerobot.
  • mechatronic devices known as factors 4 can generate force and vibration proportional to the signals detected by the tactile sensor 1 .
  • Tactors can be worn on the body or appendages or included inside a socket of a prosthetic limb to deliver these stimuli.
  • force tactors may be low-profile linear actuators or solenoids that press into the operators skin at greater depth as forces are increased.
  • vibration tactors may be low-cost vibrators, such as employed in mobile telephones, that increase the amplitude and/or frequency of vibration in proportion to the applied voltage.
  • Other methods for conveying tactile feedback such as direct electrical stimulation of the nervous system, visual displays, and audible sounds, may be used in addition or instead.
  • electronic signals from the pressure sensor 29 can be conditioned and transmitted by conventional electronic circuitry and processed by the reflex system 3 of the robot 40 on which the tactile sensor 1 may be installed.
  • Sensor electronics inside the tactile sensor 1 can be programmed to perform some or all of the desired control functions of the reflex system 3 locally.
  • the reflex system 3 can be directly integrated into the compression plate 24, which itself may be an electronics board.
  • the signals from the pressure sensor 28 can be processed so as to reveal other useful information. If the fluid 13 is well-sealed into the tactile sensor, the steady-state pressure provides a useful indication of the contact force between fingertips and objects. Small, high frequency fluctuations in the fluid pressure may be indicative of vibrations in the skin that tend to arise when the fingertips slide over surfaces and can be used to infer the texture of the surface and to identify such slippage.
  • vibration power can be determined by computing the power of filtered pressure signals within a bandwidth, such as 10-500Hz, using analog or digital signal processing methods, such as those described by Fishel JA, Loeb GE. "Bayesian exploration for intelligent identification of textures," Front. Neurorobot, 2012a, 6; and Fishel JA, Loeb GE. "Sensing tactile microvibrations with the BioTac—
  • the amplitude and timing of such vibration signals may be an indicator of incipient and actual slippage, impacts, coefficient of friction, and textural properties of surfaces.
  • Two or more tactile sensors 1 may be positioned so as to make simultaneous contact with the same object. Time-varying signals from them can be cross-correlated to distinguish impact of a grasped object with another object from slip between the robotic gripper and the grasped object. See Heyneman B, Cutkosky MR. "Biologically inspired tactile classification of object-hand and object- world interactions," 2012, pp. 167-73. All of this information derived from the fluid pressure may be used as feedback to the controller to cause automatic
  • the mechanical properties of the skin 14 can be optimized to obtain either sustained response to constant pressure or sensitivity, as illustrated in FIGS. 5A-5C.
  • Making skin 14 thicker and/or in multiple layers can reduce the leaking of the fluid 13, but may increase the rigidity and therefore reduce the sensitivity and compliance of the device.
  • FIGS. 5B-5C illustrate an example of the response of two variants of skin 14: traces A are signals from an airtight tactile sensor 1 , and traces B are from a tactile sensor 1 with a leaky skin 14.
  • the leaky skin 14 acts may act like a high pass filter, producing transient positive and negative signals at the onset and offset of contact, respectively. Absolute force could be recovered by taking the integral of the signal, but this may result in errors due to drift.
  • FIGS. 6 and 7A-7C provide an example of how contact sensitivity and compliant sensors can be used to reduce contact forces.
  • FIG. 6 is an example of a model of a robot limb segment 44 consisting of an equivalent mass m plus stiffness Krobot, and elastomeric pad 10 with stiffness Kskin, colliding with a rigid object at initial velocity v.
  • FIG. 7A is the velocity of the robot limb segment 44 vs. time during the collision and loss of contact
  • FIG. 7B is the displacement of the robot limb segment 44 vs. time during the collision and loss of contact
  • FIG. 7C is the force exerted by the robot limb segment 44 vs. time during the collision and loss of contact.
  • Four simulations are provided: a high-velocity collision with no tactile sensor (fn), a high-velocity collision equipped with a tactile sensor and reflexive ability to stop the robot once collision is detected (ft), a low-velocity collision with no tactile sensor (sn), and a low-velocity collision equipped with a tactile sensor and reflexive ability to stop the robot once a collision is detected (st).
  • Region I represents the time before collision, with collision occurring at the end of region I.
  • a robot limb segment 44 without compliant tactile sensors (cases fn and sn) quickly decelerates on impact due to the high stiffness of the robot limb segment 44 colliding with a rigid object, rapidly producing large collision forces that scale proportionally with increased initial velocity.
  • compliant tactile sensors 1 By equipping robot limb segments 44 with compliant tactile sensors 1 , collision can be detected, allowing for robotic actuator 42 to safely decelerate the robot limb segment 44 as the elastomeric pad 10 absorbs energy, thereby greatly reducing collision forces.
  • this can be embodied by permitting the reflex system 3 to identify when contact with tactile sensors 1 has occurred to modify the motor commands to the robotic actuators 42, causing them to brake or reverse the motion of limb segments 44.
  • a robot limb segment 44 with such tactile sensors 1 and a reflex system 3 permit for safe deceleration and reduced contact forces.
  • region II defines the response time of the reflex system 3 and robotic actuator 42 to begin deceleration once contact of the elastomeric pad 10 of the tactile sensor 1 attached to robot limb segment 44 is detected. During this period, the elastomeric pad 10 may begin compressing; however, the robotic actuators 42 may not yet have begun decelerating the robot limb segment 44.
  • Region III defines the period in which the robotic actuator 42 has begun decelerating the robot limb segment 44 and the elastomeric pad 10 is
  • region IV defines the period in which the elastomeric pad 10 is completely compressed, and the stiffness of the robot limb segment 44 may dominate the collision forces, rapidly decelerating the motion and producing a rapid increase in force (as in cases fn and sn).
  • the net collision force may be substantially reduced.
  • a thicker elastomeric pad 10 may be used to increase the amount of time before such a collision in which the physics are dominated by the stiffness of the robot limb segment 44.
  • the maximum velocity of the robot limb segments 44 can be limited by the controller 2 and/or the reflex system 3 to prevent such high velocities that cannot be stopped safely from ever occurring.
  • Information on the contact state can be used to provide a fragile grasp for both autonomous robotics and prosthetic hands.
  • contact information from tactile sensors 1 can be used to modulate the control signals driving the robotic actuators 42 by producing an inhibitory reflex in the reflex system 3.
  • the relationship between the command input from the user and the actual voltage of the control signal(s) transmitted to actuator(s) 42 can be described as a gain.
  • actuator 42 Upon detecting contact between one or more tactile sensors and an object to be grasped, this gain can be reduced.
  • actuator 42 is a DC motor or similar design, the sudden decrease in its control signal may result in actuator 42 stopping its motion when its torque reaches the stalling torque for the new value of the control signal, which torque may reflect the sum of frictional forces internal to its torque transmission linkage plus external force applied to the grasped object.
  • the command to actuators 42 of a robotic hand can originate from autonomous software control, from a telerobotic master who could be operating a controller, or from a prosthesis user controlling a prosthetic hand using a conventional myoelectric control as described in the above-cited reference Matulevich et al., 2013.
  • a reflex system 3 can automatically reduce the command to one or more actuators 42 by a fixed gain, permitting for automatic proportional control of stalling force regardless of when contact is made. Reducing the gain through the reflex system 3, rather than setting the gain to zero or reversing it, may permit the controller 2 to produce larger forces if desired by simply sending larger commands to the actuators 42. Such an approach may permit the controller 2 to maintain control over the behavior of the robot 40 and its actuators 42 while taking advantage of collision- mitigating reflexes.
  • the use of a short delay (either naturally occurring or deliberately added to the control loop) in this reflexive behavior when contact is detected may also permit the controller 2 to have more precise control over stopping force as described below.
  • FIGS. 8A-8D depict how contact sensitivity can be used to intelligently adjust for misalignment in a robotic grasp.
  • each finger can be stopped when contact is made, permitting for delicate grasp that does not move the object, which could cause it to become unstable and potentially tip over.
  • the recognized grasp orientation may be undesirable, as detected by position or force sensors in the articulations and/or actuators of the hand when all of the fingers are stopped, as exampled in Figure 8F-J.
  • the grasper can make use of the information to reorient the grasper to a more preferred location and to re-grasp the object.
  • the grasping can make use of the information to reorient the grasper to a more preferred location and to re-grasp the object.
  • the tactile sensors 1 may be used on robotic grippers that are equipped with larger numbers of independently positionable contact surfaces that enable more complex repositioning of objects within the gripper by
  • Relational terms such as “first” and “second” and the like may be used solely to distinguish one entity or action from another, without necessarily requiring or implying any actual relationship or order between them.
  • the terms “comprises,” “comprising,” and any other variation thereof when used in connection with a list of elements in the specification or claims are intended to indicate that the list is not exclusive and that other elements may be included.
  • an element preceded by an “a” or an “an” does not, without further constraints, preclude the existence of additional elements of the identical type.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Manipulator (AREA)

Abstract

L'invention concerne un capteur tactile conforme pouvant comprendre un matériau de type éponge, une enveloppe flexible et un capteur de pression de fluide. L'enveloppe flexible peut avoir une forme, absorber du fluide, se comprimer en réponse à l'application d'une force au matériau de type éponge, et se décomprimer et revenir à sa forme originelle lorsque la force est retirée. L'enveloppe flexible peut couvrir une surface externe du matériau de type éponge. Le capteur de pression de fluide peut détecter des changements dans la pression d'un fluide qui est à l'intérieur du matériau de type éponge provoquée par une force appliquée à l'enveloppe flexible. Un système robotique peut comprendre un bras robotique mobile, un capteur tactile conforme sur le bras robotique mobile qui détecte un contact entre le capteur tactile conforme et un objet pendant un déplacement du bras robotique mobile et qui amortit l'effet de ce contact, et un système réflexe qui amène le bras robotique mobile à se déplacer en réponse à des ordres.
EP13766791.1A 2012-09-11 2013-09-09 Capteur tactile conforme doté d'un matériau de type éponge rempli de fluide Withdrawn EP2895834A1 (fr)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
US201261699561P 2012-09-11 2012-09-11
US201361814392P 2013-04-22 2013-04-22
PCT/US2013/058809 WO2014043037A1 (fr) 2012-09-11 2013-09-09 Capteur tactile conforme doté d'un matériau de type éponge rempli de fluide

Publications (1)

Publication Number Publication Date
EP2895834A1 true EP2895834A1 (fr) 2015-07-22

Family

ID=49237643

Family Applications (1)

Application Number Title Priority Date Filing Date
EP13766791.1A Withdrawn EP2895834A1 (fr) 2012-09-11 2013-09-09 Capteur tactile conforme doté d'un matériau de type éponge rempli de fluide

Country Status (3)

Country Link
US (1) US9080918B2 (fr)
EP (1) EP2895834A1 (fr)
WO (1) WO2014043037A1 (fr)

Families Citing this family (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9625333B2 (en) 2013-03-15 2017-04-18 President And Fellows Of Harvard College Tactile sensor
EP3038699A4 (fr) 2013-08-27 2017-03-29 Halo Neuro, Inc. Système d'électrode pour stimulation électrique
US9782585B2 (en) 2013-08-27 2017-10-10 Halo Neuro, Inc. Method and system for providing electrical stimulation to a user
US9486618B2 (en) 2013-08-27 2016-11-08 Halo Neuro, Inc. Electrode system for electrical stimulation
US9630005B2 (en) 2013-08-27 2017-04-25 Halo Neuro, Inc. Method and system for providing electrical stimulation to a user
WO2017075086A1 (fr) 2015-10-26 2017-05-04 Halo Neuro, Inc. Système et procédé de positionnement d'électrode
EP3413966A4 (fr) 2016-02-08 2019-11-27 Halo Neuro, Inc. Procédé et système permettant d'améliorer la fourniture d'une stimulation électrique
WO2018005144A2 (fr) 2016-06-17 2018-01-04 Cornell University Actionneurs à base de mousse
US10485443B2 (en) 2016-06-20 2019-11-26 Halo Neuro, Inc. Electrical interface system
TWI615741B (zh) * 2017-01-06 2018-02-21 致伸科技股份有限公司 具有氣壓感測式按鍵之滑鼠
US10512770B2 (en) 2017-03-08 2019-12-24 Halo Neuro, Inc. System for electrical stimulation
US10496209B2 (en) * 2017-03-31 2019-12-03 Apple Inc. Pressure-based force and touch sensing
US10507324B2 (en) 2017-11-17 2019-12-17 Halo Neuro, Inc. System and method for individualizing modulation
US20190391651A1 (en) * 2018-06-20 2019-12-26 Mayu, Inc. Flexible and tactile pressure sensitive switch sensors
JP7107017B2 (ja) * 2018-06-21 2022-07-27 カシオ計算機株式会社 ロボット、ロボットの制御方法及びプログラム
JP6913906B2 (ja) * 2018-10-24 2021-08-04 株式会社三重ロボット外装技術研究所 接触検出装置
GB2580164A (en) * 2018-12-21 2020-07-15 Imperial College Sci Tech & Medicine A sensor
JP7093080B2 (ja) * 2019-01-15 2022-06-29 豊田合成株式会社 触感センサ
US11847311B2 (en) * 2019-05-22 2023-12-19 Apple Inc. Characterization of a venting state or other system parameter that affects the characterization of a force applied to a device
EP3975826B1 (fr) * 2019-06-03 2024-06-26 3M Innovative Properties Company Systèmes d'évaluation de déficiences de perception haptique et d'anomalies de motricité
JP2021091060A (ja) * 2019-12-12 2021-06-17 セイコーエプソン株式会社 制御方法およびロボットシステム
US11408788B2 (en) * 2020-03-31 2022-08-09 Toyota Research Institute, Inc. Variable geometry and stiffness control for fluid filled sensor
US11668616B2 (en) * 2021-03-30 2023-06-06 The Board Of Trustees Of The Leland Stanford Junior University Stretchable tactile sleeve for robotic extremities
US11993498B2 (en) 2021-10-28 2024-05-28 Toyota Research Institute, Inc. Structures and sensor assemblies having engagement structures for securing a compliant substrate assembly
US11867574B1 (en) 2023-05-23 2024-01-09 Sanctuary Cognitive Systems Corporation Fluidic tactile sensor

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090031825A1 (en) * 2007-07-31 2009-02-05 Takeo Kishida Detecting device
US20120098789A1 (en) * 2010-10-20 2012-04-26 Craig Michael Ciesla User Interface System

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4306148A (en) 1980-01-21 1981-12-15 General Electric Company Tactile sensor
US7555346B1 (en) 1999-01-07 2009-06-30 Boston Scientific Neuromodulation Corporation Implantable pulse generator having current steering means
US6529122B1 (en) 1999-12-10 2003-03-04 Siemens Technology-To-Business Center, Llc Tactile sensor apparatus and methods
US6443509B1 (en) 2000-03-21 2002-09-03 Friendly Robotics Ltd. Tactile sensor
US8181540B2 (en) 2006-03-28 2012-05-22 University Of Southern California Measurement of sliding friction-induced vibrations for biomimetic tactile sensing
EP2010868B1 (fr) 2006-03-28 2017-01-18 University of Southern California Capteur tactile biomimetique
US7878075B2 (en) 2007-05-18 2011-02-01 University Of Southern California Biomimetic tactile sensor for control of grip
WO2008032661A1 (fr) * 2006-09-12 2008-03-20 National Institute Of Advanced Industrial Science And Technology Procédé de mesure de valeur de distribution et système de mesure utilisant un capteur de valeur de distribution pour celui-ci
US7701202B2 (en) * 2006-11-02 2010-04-20 Massachusetts Institute Of Technology Compliant tactile sensor that delivers a force vector
US20110190805A1 (en) * 2010-02-01 2011-08-04 Reinhard Carl Huber Intraurethral Retainer and Keeper For Facilitating Penile Fixation and/or Occlusion

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090031825A1 (en) * 2007-07-31 2009-02-05 Takeo Kishida Detecting device
US20120098789A1 (en) * 2010-10-20 2012-04-26 Craig Michael Ciesla User Interface System

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of WO2014043037A1 *

Also Published As

Publication number Publication date
WO2014043037A1 (fr) 2014-03-20
US9080918B2 (en) 2015-07-14
US20140069212A1 (en) 2014-03-13

Similar Documents

Publication Publication Date Title
US9080918B2 (en) Compliant tactile sensor with fluid-filled, sponge-like material
Kim et al. Soft inflatable sensing modules for safe and interactive robots
KR102477836B1 (ko) 마스터 로봇 시스템용 저항 기반 관절 구속
Kappassov et al. Tactile sensing in dexterous robot hands
AU2015305311B2 (en) Sensors for soft robots and soft actuators
Krahn et al. A soft-touch gripper for grasping delicate objects
US10551923B2 (en) Haptic glove as a wearable force feedback user interface
JP5187856B2 (ja) 触覚センサ
Wang et al. Mechanoreception for soft robots via intuitive body cues
US8181540B2 (en) Measurement of sliding friction-induced vibrations for biomimetic tactile sensing
US8272278B2 (en) Enhancements to improve the function of a biomimetic tactile sensor
US9857245B2 (en) Soft-body deformation and force sensing
BR102017008927B1 (pt) Dispositivo automatizado
BR102017008920B1 (pt) Dispositivo automatizado com uma estrutura móvel, em particular, um robô e cobertura sensorizada para cobrir tal estrutura
Kappassov et al. Semi-anthropomorphic 3D printed multigrasp hand for industrial and service robots
Vogt et al. A soft multi-axis force sensor
WO2009124209A1 (fr) Mesure fiable des vibrations induites par le frottement de glissement pour la détection tactile biomimétique
Suh et al. Soft pneumatic actuator skin with embedded sensors
Cheng et al. Soft fabric-based pneumatic sensor for bending angles and contact force detection
Tawk et al. Position control of a 3D printed soft finger with integrated soft pneumatic sensing chambers
Gruebele et al. A stretchable tactile sleeve for reaching into cluttered spaces
Ma et al. Sensor embedded soft fingertip for precise manipulation and softness recognition
Nishikawa et al. Development of pneumatically driven hand capable of grasping flexible objects
Nakanishi et al. On agreement between tactile sensing and rendering
Karatoa et al. Perception of Approaching Objects in Bilateral Control Using Proximity Sensor

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20150326

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

AX Request for extension of the european patent

Extension state: BA ME

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: SYNTOUCH, LLC

DAX Request for extension of the european patent (deleted)
17Q First examination report despatched

Effective date: 20170220

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN

18D Application deemed to be withdrawn

Effective date: 20170704